In responses to the April 1 2008 Climate Science weblog, I e-mailed to Josh Willis to both alert him to the weblog and follow up on his comments about upper ocean heat trends. With his permission I am posting our e-mail exchanges below. These adds to our knowledge of this critically important climate metric of global warming and cooling.
E-mail Inquiry
Hi Josh,
I have read your op-ed on the weblog http://network.nationalpost
It seems you have developed a difference of interpretation of your data from what you had e-mailed previously. The Argo data indicates that if global warming occurred over the last 4 years, it was in the deeper ocean and is thus not available in the short term to the atmosphere (in fact, if it is there, it likely more diffused and therefore could only enter the atmosphere slowly if at all), or it has exited into space. The other heat stores in the climate system are too small (and the atmosphere has clearly not warmed over the last few years). Global sea ice cover is actually above average at present (the Antarctic sea ice is at a near record level). The continued sea level rise indicates to me that the heat is in the deeper ocean (which is not predicted by the models).
There is also no unrealized heat in the system. This is a fallacy of using temperature trends as the surrogate for heat trends as I have reported on my weblog. The GCM models also do not report on 4 year time periods without any warming for the current decades.
I will comment further in my weblog but I am perplexed by your dismissal of the actual significance of your results. It appears you have jumped on the advocacy bandwagon with your last paragraph.
Best Regards
Roger
First E-mail Reply
Hi Roger,
Thanks for the email. Attached is a plot showing the upper 750 m ocean heat content from the GFDL CM2.1 coupled climate model. The blue line is the 12-month running mean heat content from one run and the green line shows the monthly-mean heat content for the same run, but with the control run subtracted. Subtracting the control run essentially takes out the seasonal cycle. Even apart from the period after the eruption of Mt. Pinatubo in the early 1990s there are several multi-year periods when OHCA is flat.
Also, I believe that after all the dust settles on the XBT bias corrections, the 50 year observational record of OHCA will have still have several multi-year periods with little or no warming.
I do agree with you that heat content is a much better measure of forcing than surface temperature. I also agree that the cause for this most recent flat period is not understood and figuring it out is important for understanding climate change. Furthermore, if it is some type of natural variability (which seems reasonable to me), it is important for the IPCC models to simulate it correctly, especially if there is an effort to move toward decadal projections instead of just the centennial projections that they now produce.
That said, it does seem very likely to me that the 50-year trend in ocean heat content is part of the climate’s ongoing response to greenhouse forcing. If that is the case, I think it is pretty easy to argue that the oceans and surface temperature will be warmer 50 years from now than they are today, and this is the reason for my comment in the National Post.
As for the sea level rise discrepancy, I do not think that it is possible for enough heat to have been absorbed by the deep oceans to fully explain the difference between the altimeter and the upper ocean thermal expansion we reported in the JGR paper. There have been many repeat hydrographic sections that show warming in the deep ocean, but inventories of these suggest that a few tenths of a mm/yr is the most that could be accounted for by this. This points to some kind of systematic error in one of the three observing systems, which is not entirely unlikely.
Cheers,
Josh
My Reply
Hi Josh, Thanks for the further feedback. Can you send the figure with just the forced run on it? It would be clearer to just have that plot.
On the reasons for sea level rise, I asked the question about rises in the ocean sea bed when I was at the University of Texas at Austin last year, and they told me it was too small of an effect. This is certainly not my expertise, but I would be interested in further discussion on the uncertainties of this measurement for decadal and yearly time scales.
On model performance on ENSOs, PDO and the NAO, they certainly do a poor job on seasonal time scales (as exemplified for this past winter), and I would be interested in any papers that have examined these specific circulation indices on longer time periods. I know that some of the models produce ENSO events but do not know how realistic in frequency or magnitude they are.
There was a talk on this by De-Zheng Sun of CIRES in my class last year;
http://cires.colorado.edu
where he found the models, when run in a coupled atmosphere-ocean mode, over predicted ocean surface warming.
Roger
P.S. On posting your comments, can I also cut text out of your latest e-mail to add? I will send the final draft to you first before posting to make sure it accurately conveys your viewpoint.
Second E-Mail Reply
Hi Roger,
New plot attached. As a rule, you are welcome to quote my emails whenever you want on the blog. I’ll let you know if there’s something I don’t want you to share. :)
The issue of the sea bed moving is one that has been given some thought by the geodesists, and in our paper we make an adjustment 0.3 mm/yr to the altimeter for this. All of the people I’ve talked to have told me that this should be fairly linear over time scales of a few hundred years, but I don’t know that anyone has looked into short time scale variability specifically.
Cheers, Josh
Third E-Mail Reply
Hi Roger, Yes, you are more than welcome to post the comments and plot. Never mind the forced-control line. That’s leftover from my habit of looking at the deep ocean, which has a significant drift.I haven’t looked at the radiative imbalance in this model, but there is a pretty large amount of heat entering the deep ocean. I’m not sure how realistic this is (that’s one of the things I’m looking into), but getting that right is important for getting the overall radiative imbalance and I suspect that lots of the IPCC models do a bad job in the deep ocean. You are right that this model only has very short periods (4 years at most) and I suspect that they have something to do with El Nino in this particular model. That’s also something I’m looking into. But of course in the real ocean there are other quasi-global scale ocean phenomena, like the PDO, and I don’t have any idea if any climate models get that right.
Cheers, Josh
_______________________________________________________
Climate Science Conclusion
- The ocean heat content remains the most appropriate climate metric to monitor global warming (or cooling).
- The models (of which the figure presented above is a representative example)show a more-or-less monotonic increase in upper ocean heat content in coming years.
- Any leveling off of this increase is short lived in the models and, at most 4 years in the example presented above.
- The recent (last 4 years) upper ocean observations show no warming.
- If heat is entering the deep ocean, it is, therefore, not contributing to an increase in sea surface temperature. This necessarily reduces the amount of water vapor evaporated into the atmosphere from what would occur if the heat were in the upper ocean.
- The assessment of upper ocean heat content provides an ideal quantitative comparision tool with the models, using the approach presented on Prometheus.